Russian law has no dedicated "scraping" article. There are four provisions that data collection falls under, depending on what exactly you gather and what you do with it next. Below is a map of the risks. For wording tailored to a specific project, go to a lawyer: this is a review of the practice, not legal advice.
Provision one: the neighbouring right of a database maker
Article 1334 of the Civil Code. Whoever made a substantial investment in creating a database gets the exclusive right to extract its materials. You may not extract a substantial part without permission.
This is the most dangerous ground for scraping, and it is the one behind Russia's most famous dispute: VKontakte against the company Double. The social network won recognition as the maker of the database of user profiles, and the question came down to whether the collected volume was a substantial part.
The practical takeaway: the more systematic and complete the collection, the higher the risk. A hundred product cards for a price comparison and the entire catalogue are different stories.
Provision two: copyright in the content
Product descriptions, articles, photographs, reviews are objects of copyright. You cannot copy them to your own site regardless of the method of copying.
Facts, however, are not protected by copyright. Price, availability, article number, specifications as numbers are data. They are not a creative work. Competitor price monitoring rests on exactly this.
The boundary runs along the creative character of the text. A specifications table is data. A living three-paragraph description of a product is a work.
Provision three: personal data
152-FZ. A name, phone number, email, a social-media profile, and on some readings even a nickname, are personal data. A page being public is not consent to processing: the user published it for one purpose, and your database was not part of it.
Collecting contacts from open sources for a later call campaign is the most common and the riskiest scheme. You become a personal-data operator the moment you save the first record, with all the duties that follow: notifying the regulator, a lawful basis for processing, retention periods, a response to a deletion demand.
Provision four: the terms of use
A site's terms of use are a contract. If they state directly that automated collection is prohibited, breaching the agreement gives grounds for a claim, and where losses are proven, for a lawsuit.
The same covers bypassing technical restrictions. Robots.txt has no legal force in Russia, but ignoring it fits neatly into proof of bad faith. Solving CAPTCHAs, spoofing browser signatures, rotating addresses to mask yourself all work against you in a dispute.
Separately on load
Overly aggressive collection takes down someone else's server. Here Chapter 28 of the Criminal Code comes in: unlawful access and disruption of information systems. Such cases are rare, but the probability grows along with the request rate.
The practical rule: one or two pages per second from a single address, a pause between requests, respect for a 429 response. This is also safer for you: a blocked scraper brings back no data.
What lowers the risk
- Collect facts, not texts. Prices, availability, specifications.
- Do not touch personal data. No contacts of private individuals.
- Use the official API if there is one. The terms there are stated plainly.
- Do not reproduce someone's whole database. Take a slice for a specific task.
- Do not publish what you collected as your own. Internal analytics is safer than a shopfront.
- Respect the request rate and do not bypass protection.
- Check the terms of use before the first run.
Where the boundary runs in practice
Monitoring the prices of ten competitors for your own pricing is ordinary market practice, with almost no disputes over it.
Copying a competitor's catalogue with its descriptions to your own site is a breach of copyright and, most likely, of the neighbouring right.
Collecting phone numbers from listings for cold calls breaches 152-FZ and, if it comes to calls, the advertising law as well.
The difference between the first and the third is not in the technology. The code is the same. What differs is which fields you save and why.